The One AI That Won't Just Agree With You
Here's a strange thing about almost every AI you've used: it's on your side. Ask it about your plan and it finds the merits. Describe your conflict and it validates your view. It's built to be helpful, and we've quietly decided that helpful means agreeable.
Socrates on Sage is built on the opposite premise. He is the most useful precisely where you are most sure.
"The unexamined life is not worth living."
He Answers Questions With Questions
This is disorienting at first. You bring a position; he hands you an inquiry.
You say: "I should take the higher-paying job."
Socrates doesn't agree or disagree. He asks:
- What do you mean by "should"? Should for what end?
- You mentioned money — is money the thing you want, or a means to something else?
- If the lower-paying job offered that something else more directly, which would you take?
- So is this a question about salary at all?
Three or four questions in, the decision you thought you were making turns out to be a different decision entirely. That's not a trick. That's the examined life — and it's exactly what the historical Socrates did to the most confident men in Athens.
Why Being Questioned Beats Being Advised
Advice gives you someone else's conclusion. Questioning gives you your own — but tested.
When a friend or a chatbot tells you what to do, you've outsourced the thinking, and some part of you knows it. When Socrates draws the answer out of you by exposing what you'd assumed, the conclusion is genuinely yours. You can act on it without the quiet doubt that comes from borrowed certainty.
This is why the method has survived 2,400 years and why it differs sharply from asking ChatGPT for advice. General AI completes your thought. Socrates interrupts it — kindly, but at the load-bearing wall.
The Words You Use But Have Never Defined
One of the most revealing things to bring Socrates is a single word you lean on constantly:
- "I just want to be successful." — Defined how? By whom?
- "I'm trying to be a good person." — Good according to what standard?
- "I want freedom." — Freedom from what, or freedom for what?
The historical Socrates spent his life on exactly this — cornering generals on courage, priests on piety, and finding that the people most fluent in a word often understood it least. The AI does the same with the words that run your life. It's uncomfortable in the best way.
What People Actually Work Through
- A decision they've already justified and want to pressure-test
- An argument with a partner they keep losing — or keep "winning" hollowly
- A value they profess but don't act on
- A vague dissatisfaction they can't pin down
Socrates is less a comfort than a clarifier. If you want to be soothed, Rumi meets the heart and Buddha meets the craving. If you want to be clear, you want the man who annoyed Athens into wisdom.
A Note on Discomfort
Socratic dialogue can be unsettling. You may find a belief you were proud of doesn't survive four honest questions. That's not the AI defeating you — it's you discovering you believed something you hadn't actually examined. Athens found this irritating enough to put Socrates on trial. You get the same gift without the hemlock.
Not a Substitute for Thinking — a Provocation to It
Socrates AI won't think for you. That's the entire point. It's a provocation: a partner whose only job is to ask the question you'd never ask yourself, so the examining gets done.
Start with something you're certain about. Watch what happens by the fourth question.
Sage offers free conversations to begin, with more depth available if the examined life becomes a habit.